WO2011116157A1

WO2011116157A1 - Catalytic fixed bed reactor for producing ethylene oxide by partial oxidation of ethylene

Info

Publication number: WO2011116157A1
Application number: PCT/US2011/028763
Authority: WO
Inventors: Max M. Tirtowidjojo; Christina Zarth; Subrata Sen; Pramod D. Patil
Original assignee: Dow Technology Investments Llc
Priority date: 2010-03-17
Filing date: 2011-03-17
Publication date: 2011-09-22
Also published as: TWI496618B; US20130144073A1; US10010856B2; CA2793500A1; CN102905782A; US9073034B2; CA2793500C; EP2547434A1; CN102905782B; EP2547434B1; US20150336064A1; TW201143892A

Abstract

A method and a reaction vessel for producing gaseous ethylene oxide from partial oxidation of hydrocarbon using a heterogeneous catalyst fixed bed comprising: - introducing gaseous ethylene, oxygen, ballast gas, and catalyst promoter into a reaction vessel having a catalyst bed with a length such that an outflow area and an inflow area over said catalyst bed length in between the reactor outflow and inflow has an absolute ratio difference less than or equal to 1.3 m; said catalyst having a selectivity greater than 80%; - circulating a heat transfer fluid through at least one coolant heat exchanger in said shell interior; said heat exchanger having a coolant flow cross sectional area ratio to cooling surface area less than 1; - flowing said gaseous ethylene, oxygen, ballast gas and catalyst promoter over said catalyst and through an outlet zone in said reactor vessel, said zone configured with an average residence time of less than or equal to 4 seconds from the catalyst bed to the heat exchanger to quench said process to create gaseous ethylene oxide product.

Description

CATALYTIC FIXED BED REACTOR FOR PRODUCING ETHYLENE OXIDE BY PARTIAL OXIDATION OF ETHYLENE

TECHNICAL FIELD

[0001] Currently, reactors for producing ethylene oxide (EO) by partial oxidation of ethylene typically make use of a single conventional fixed-bed shell-tube exchanger (FB) where the catalytic reaction occurs inside the tubes. The fabrication of this type of reactor has reached engineering and transportation limitations due to weight and size factors. It is typical in a conventional design to have 8,000-14,000 tubes with up to 2" internal diameter tubes (Dt) arranged in the shell diameter (Ds) with a Ds of 6-9 meters (M) and tube sheets approaching 1.0 ft to 2.0 ft in thickness. A reactor with an even larger Dt and Ds can theoretically be used to provide a more economically advantageous process given the continually advancing

catalyst formulations that are improving average selectivity of 80% to 95% during such a reactor's life time. There is a need for a reactor configuration as an alternative platform for EO catalyst with efficiency higher than 80% that is lower in weight and provides lower pressure drop across the reactor and thus provides higher return on capital investment due to lower operating and capital cost as compared to using a conventional FB reactor.

BRIEF SUMMARY

[0002] In one embodiment, the disclosure relates to a reaction vessel for production of

alkylene oxide(s) from partial oxidation of hydrocarbon using a high efficiency

heterogeneous catalyst in a fixed bed enclosed within a reaction vessel shell. The reaction vessel may comprise a shell having a length and a volume that defines a catalyst bed shape having a length such that an out flow area and an in flow area over the catalyst bed length in between the out flow and in flow has an absolute ratio difference less than or equal to about 1.3 M anywhere in the reactor bed. The catalyst bed defines a process side having a

selectivity greater than about 80%, and the catalyst bed has a length less than the shell length and a width that defines a volume less than the shell volume. The reaction vessel further includes a fixed bed outlet zone configured with average residence time less than or equal to about 4 seconds of the gaseous product flow from the catalyst bed over the heat exchanger to quench the undesirable side reactions involving the alkylene oxide product. The vessel also includes at least one fluid coolant enclosure heat exchanger in the vessel interior with an outside surface and an inside surface. The coolant enclosure outside surface is in contact with the catalyst bed. The coolant enclosure has an inlet and an outlet for the flow of heat transfer fluid therethrough. The coolant enclosure may further define a cooling surface area with the coolant flow cross sectional area ratio to cooling surface area much less than about 1 and where pressure in the coolant side may be higher than pressure on the process side.

[0003] In another embodiment, the disclosure relates to at least one method for producing ethylene oxide from partial oxidation of ethylene using a high efficiency ethylene oxide catalyst in a fixed bed enclosed within a shell of a reaction vessel. In one embodiment, the method may comprise introducing a sufficient amount of gaseous ethylene, oxygen, ballast gases that include, but are not limited to, methane, inert gases such as N₂, He, Ar and any other inert gas, and at least one catalyst promoter such as, but not limited to, nitric oxide, vinyl chloride, ethyl chloride, and others, into an in flow of the reaction vessel and flowing the ethylene, oxygen, ballast gas and promoters over an ethylene oxide (EO) catalyst bed that provides a selectivity to EO of greater than about 80%. The reaction vessel may comprise a shell having a length and a volume that defines a catalyst bed shape having a length such that an out flow area and an in flow area over the catalyst bed length in between the out flow and in flow has an absolute ratio difference less than or equal to about 1.3 M anywhere in the reactor bed. The method further includes circulating a heat transfer fluid within a coolant enclosure contained within the reaction vessel catalyst bed. The coolant enclosure defines a coolant side, and the coolant side may have a greater pressure than the process side. The coolant enclosure has an outside surface in contact with the catalyst bed, and has an inlet and an outlet for the circulation of heat transfer fluid therethrough. Generally, the coolant enclosure defines a cooling surface area with a coolant flow cross sectional area ratio to cooling surface area much less than about 1. The reaction vessel further includes a fixed bed outlet zone configured with an average residence time less than or equal to 4 seconds of gaseous product flow from the outlet of the catalyst bed over the heat exchanger to quench any undesirable side reactions involving the ethylene oxide product.

BRIEF DESCRIPTION OF THE DRAWINGS

[0004] FIG. 1 is a schematic representation of a reaction vessel according to at least one embodiment;

[0005] FIG. 1A is a cross sectional view of the heat transfer fluid enclosure of FIG. 1; [0006] FIG. 2 A is an evaluation of a catalyst-in- shell side reactor design with cross flow (XCSA) for a low selectivity catalyst with a GHSV of 5631 1/hr;

[0007] FIG. 2B is an evaluation of a catalyst-in- shell side reactor design with cross flow (XCSA) for a low selectivity catalyst with a GHSV of 7525 1/hr;

[0008] FIG. 2C is a plot of heat transfer area to catalyst volume ratio φ as function of reactor bed length for XCSA reactor designs with various coolant tube diameters and various coolant temperatures yielding different work rates for a low selectivity catalyst;

[0009] FIG. 2D is a plot of heat transfer area to catalyst volume ratio φ as function of reactor bed length for XCSA reactor designs at different GHSV values for a low selectivity catalyst;

[0010] FIG. 2E is a plot of heat transfer area to catalyst volume ratio φ as function of reactor bed length for XCSA reactor designs with various coolant tube diameters and various coolant temperatures yielding different work rate for a low selectivity catalyst;

[0011] FIG. 2F is a plot of heat transfer area to catalyst volume ratio as function of reactor bed length for XCSA reactor designs with various tube diameters and various coolant temperatures yielding different work rate for a low selectivity catalyst;

[0012] FIG. 2G is a plot of heat transfer area to catalyst volume ratio as function of reactor bed length for XCSA reactor designs with various tube diameters and various coolant temperatures yielding different work rate for a low selectivity catalyst;

[0013] FIG. 2H is a plot of NPV savings of a feasible XCSA design with coolant tube as a function of bed length, as compared to an STR case with various heat transfer area ratios to catalyst volumes (or tube ID) for a low selectivity catalyst;

[0014] FIG. 3A is an evaluation of catalyst-in-shell side reactor design with cross flow (XCSA) for a high selectivity catalyst with a GHSV of 6652 1/hr;

[0015] FIG. 3B is an evaluation of catalyst-in-shell side reactor design with cross flow (XCSA) for a high selectivity catalyst with a GHSV of 8500 1/hr; [0016] FIG. 3C is a plot of heat transfer area to catalyst volume ratio for high selectivity catalyst as a function of reactor bed length for XCSA reactor designs with various tube diameters;

[0017] FIG. 3D is a plot of heat transfer area to catalyst volume ratio for a high selectivity catalyst as function of reactor bed length for XCSA reactor designs having different GHSV values;

[0018] FIG. 3E is a plot of the heat transfer area to catalyst volume ratio as function of reactor bed length for XCSA reactor designs with various coolant temperatures yielding different work rate for a high selectivity catalyst;

[0019] FIG. 3F is a plot of heat transfer area to catalyst volume ratio as function of reactor bed length for XCSA reactor designs with various coolant temperatures yielding different work rate for a high selectivity catalyst;

[0020] FIG. 3G is a plot of NPV savings of feasible XCSA designs with coolant tubes OD of 0.75" as a function of bed length, as compared to an STR case with various heat transfer area ratio to catalyst volume (or tube ID) for a high selectivity catalyst;

[0021] FIG. 4 is a schematic of radial flow reactor and cone shaped catalyst bed reactor designs;

[0022] FIG. 5 A is an evaluation of catalyst-in- shell side axial flow designs with flow parallel to the coolant carrier (CSA) for low selectivity catalyst as compared to an STR case;

[0023] FIG. 5B is an evaluation of catalyst-in-shell side axial flow designs with flow parallel to the coolant carrier (CSA) for low selectivity catalyst as compared to an STR case;

[0024] FIG. 6 A is an evaluation of catalyst-in-shell side axial flow designs with flow parallel to the coolant carrier (CSA) for high selectivity catalyst as compared to an STR case;

[0025] FIG. 6B is an evaluation of catalyst-in-shell side axial flow designs with flow parallel to the coolant carrier (CSA) for high selectivity catalyst as compared to STR case; [0026] FIG. 7 shows the comparison of XCSA reactor designs and conventional reactor designs performance of various ranges of catalyst bed porosity and density for low selectivity catalyst;

[0027] FIG. 8 shows the comparison of an XCSA and conventional reactor designs performance for various ranges of catalyst bed porosity and density for a high selectivity catalyst.

DETAILED DESCRIPTION OF THE INVENTION

[0028] Turning now to the drawings wherein like numbers refer to like structures, Figure 1 is a schematic representation of a reaction vessel 10 having a shell enclosure 12 of a length and width that defines an internal space 14. The shell is a wall with an inner surface 16 and an outer surface 18, separated by an insulation layer 20. While the shell is shown schematically, those skilled in the art understand that it could be constructed in any shape desired. The shell is constructed of materials having sufficient strength to contain the internal pressures that arise as the operation of the reaction vessel as is well known in the art. The reaction vessel is further equipped with an in flow 22 for ingress of hydrocarbon or other gaseous raw materials, such as, for example ethylene, oxygen, ballast gases, and gaseous catalyst promoters, into the feed distribution device 24 and an out flow 26 for effluent gaseous products. In this regard, it is apparent to those skilled in the art that the term "ballast gases" are understood to be, but are not limited to, C0₂, CH₄, inert gasses such as N₂, Helium, Argon, or any other noble gas. Similarly, catalyst promoters may be, but are not limited to, nitric oxide (NO), especially for high selectivity catalysts, chlorides, vinyl chloride, ethyl chloride, ethane, and any other suitable gaseous promoter. The in flow and the out flow of the shell are configured to produce an exit gas velocity from about 5 ft/s (1.5m/s), to about 25 ft/s (7.6m/s), upon exiting the fixed bed reaction zone within the reactor vessel before entering an outlet pipe 31. The gaseous in flow and out flow have an absolute difference ratio of inlet and outlet flow area over the catalyst bed length from about 0.8 meters to about 1.3 meters and more preferably from about 0.9 to about 1.2 meters. In one embodiment, the shell has an interior pressure during gaseous raw material ingress and gaseous effluent product egress of less than about 350 psig.

[0029] Between the in flow and the out flow, there is a catalyst bed 28 carried within the shell selected from a high or low selectivity catalyst for the oxidation of the gaseous raw material, such as, for example ethylene, to ethylene oxide in a manner to be hereinafter described. The catalyst bed is also known as the process side of the reactor and has a length and a width and defines a volume that is less than the volume of the shell. At any point Al and A2, between the in flow and the out flow, the catalyst bed has a length LI such that between the out flow area and the in flow area over the catalyst bed length in between the out flow and the in flow, has an absolute ratio difference as expressed in (A2-A1)/L1 that is less than or equal to about 1.3 m. Proximal to the out flow is fixed bed outlet zone 30 to minimize residence time of the effluent product material after exiting the catalytic bed.

Generally, the fixed bed outlet zone is configured with an average residence time of less than or about 4 seconds for the gaseous product flow from the outlet of the catalyst bed to a heat exchanger to quench any undesired side reaction further converting alkylene oxide product to other unwanted byproducts such as C0₂, H₂0, carbon, and CH₄.

[0030] The catalyst may be selected from the group of catalysts exhibiting a selectivity toward ethylene oxide higher than about 80% under the desired reaction conditions at any point during catalyst life. The "selectivity" of the epoxidation reaction, which is synonymous with "efficiency," refers to the relative amount (as a fraction or in percent) of converted or reacted olefin that forms a particular product. For example, the "efficiency to ethylene oxide" refers to the percentage on a molar basis of converted or reacted ethylene that forms ethylene oxide.

[0031] The "heterogenous catalyst" comprises a catalytic metal and a support. The support (also known as a "carrier") may be selected from a wide range of inert support materials. Such support materials may be natural or artificial inorganic materials and they include silicon carbide, clays, pumice, zeolites, charcoal and alkaline earth metal carbonates, such as calcium carbonate. Preferred are refractory support materials, such as alumina, magnesia, zirconia and silica. The most preferred support material is a-alumina. In one exemplary embodiment, silver is deposited on the catalyst carrier as are one or more solid promoters.

[0032] There are many well-known methods of preparing supports suitable for use in ethylene oxide catalysts. Some of such methods are described in, for example, U.S. Patents 4,379,134; 4,806,518; 5,063,195; 5,384,302, U.S. Patent Application 20030162655 and the like, incorporated herein by reference. For example, an alpha-alumina support of at least 95 % purity can be prepared by compounding (mixing) the raw materials, extrusion, drying and a high temperature calcination. In this case, the starting raw materials usually include one or more alpha-alumina powder(s) with different properties, a clay-type material which may be added as binder to provide physical strength, and a burnout material (usually an organic compound) used in the mix to provide desired porosity after its removal during the calcination step. The levels of impurities in the finished carrier are determined by the purity of the raw materials used, and their degree of volatilization during the calcination step.

Common impurities may include silica, alkali and alkaline earth metal oxides and trace amounts of metal and/or non-metal-containing additives. Another method for preparing a carrier having particularly suitable properties for ethylene oxide catalyst usage comprises optionally mixing zirconium silicate with boehmite alumina (AIOOH) and/or gamma- alumina, peptizing the aluminas with a mixture containing an acidic component and halide anions (preferably fluoride anions) to provide peptized halogenated alumina, forming (for example, by extruding or pressing) the peptized halogenated alumina to provide formed peptized halogenated alumina, drying the formed peptized halogenated alumina to provide dried formed alumina, and calcining the dried formed alumina to provide pills of optionally modified alpha-alumina carrier.

[0033] The alpha-alumina carrier preferably has a specific surface area of at least about 0.5 m²/g, and more preferably, at least about 0.7 m²/g. The surface area is typically less than about 10 m²/g, and preferably, less than about 5 m²/g. The alpha- alumina carrier preferably has a pore volume of at least about 0.3 cm³/g, and more preferably, from about 0.4 cm³/g to about 1.0 cm³/g and a median pore diameter from about 1 to about 50 microns. A variety of carrier morphologies may be used, including pills, cylinders, cylinders with one or more longitudinal axial openings, chunks, tablets, pieces, pellets, rings, spheres, wagon wheels, saddle rings and toroids having star shaped inner and/or outer surfaces.

[0034] Catalysts for the production of alkylene oxide, for example, ethylene oxide or propylene oxide may be prepared with the aforementioned carriers by impregnating the carrier with a solution of one or more silver compounds, depositing the silver throughout the pores of the carrier and reducing the silver compound as is well known in the art. See for example, Liu, et al, U.S. Patent No. 6, 511 ,938 and Thorsteinson et al, U.S. Patent No. 5,187,140, incorporated herein by reference.

[0035] Examples of well-known solid promoters for catalysts used to produce ethylene oxide include compounds of potassium, rubidium, cesium, rhenium, sulfur, manganese, molybdenum, and tungsten. During the reaction to make ethylene oxide, the specific form of the promoter on the catalyst may be unknown. Examples of solid promoter compositions and their characteristics as well as methods for incorporating the promoters as part of the catalyst are described in Thorsteinson et al, U.S. Patent No. 5,187,140, particularly at columns 11 through 15, Liu, et al, U.S. Patent 6,511,938, Chou et al, U.S. Patent No. 5,504,053, Soo, et al, U.S. Patent No. 5,102, 848, Bhasin, et al, U.S. Patent Nos. 4, 916,243, 4,908,343, and 5,059,481, and Lauritzen, U.S. Patent Nos. 4,761,394, 4,766,105, 4,808,738, 4,820,675, and 4,833,261, all incorporated herein by reference. The solid promoters are generally added as chemical compounds to the catalyst prior to its use.

[0036] The catalyst may be pills having a diffusion length from about 0.02 inches to about 0.07 inches, and more preferably, from about 0.025 inches to about 0.06 inches. The pill diffusion length can be determined by the ratio of the volume of a catalyst pellet to its exterior surface available for reactant penetration and diffusion. A more detailed definition and example can be found on page 476 of "Chemical Reaction Engineering", second edition, Wiley & Sons, 1972 incorporated in its entirety by reference. If a low selectivity catalyst is used, the preferred catalyst bed should have a length greater than or equal to about 9.5 m, and if a high selectivity catalyst is used as the catalyst bed, the preferred bed should have a length greater than or equal to about 8.5 m.

[0037] The reactor vessel is further equipped with a coolant fluid enclosure heat exchanger 32 having an inlet 34 and an outlet 36 for the circulation of heat transfer fluid through the vessel in a manner that may be parallel or cross wise to the direction of gaseous raw material flow through the catalyst bed. As seen in FIG. 1 A, the coolant enclosure is generally designated as the coolant side of the reactor, and has an outer surface 38 in contact with the catalyst bed, and an inner surface 40, in contact with the heat transfer fluid. The coolant enclosure defines a coolant surface area with a heat transfer fluid flow cross sectional area ratio to the coolant surface area much less than 1. Moreover, the coolant enclosure surface area ratio to catalyst bed volume is preferably less than or equal to about 187 1/m. The heat transfer fluid may be boiling water in the coolant enclosure at a pressure of up to about 750 psig or 5170kPa gauge to maintain temperature in the catalyst bed at a temperature up to about 270°C. In addition, the pressure in the coolant side is preferably greater than the pressure on the process side of the reaction vessel. [0038] Generally, the catalyst bed has an oxidation catalyst exhibiting a selectivity greater than about 80%, and, as previously stated, the flow of gaseous raw material through the catalyst bed may be parallel or crosswise to the direction of flow of heat transfer fluids in the heat transfer fluid enclosure. Accordingly, the coolant enclosure flow may be parallel, helical, perpendicular or in any other direction to the direction of flow of the gaseous raw material through the catalyst bed.

EXAMPLES

[0039] The following examples are offered to illustrate various aspects of the present invention. Those skilled in the art understand that they are not to be construed as limiting the scope and spirit of the invention. For all Figures discussed in the Examples, "XCSA" means cross flow catalyst-in-shell reactor; "GHSV" means Gas Hourly Space Velocity, "φ" means catalyst volume ratio; "SI" means calculated sensitivity index; "ΔΡ" means pressure drop and "STR" means conventional reactor with catalyst bed inside the tube.

[0040] Table A:

[0041] The reactor conditions given in Table A apply to all of the calculations shown in the following examples, except where otherwise indicated.

Example 1

[0042] A comparison was made between a low selectivity (LS) ethylene oxide catalyst system and a high selectivity (HS) ethylene oxide catalyst system, with their relevant parameters particle diameter (Dp), porosity (ε), bed density (p_B), beginning of life (BOL) selectivity and typical end of life (EOL) selectivity. Table 1 lists some typical parameters for the low selectivity catalysts and the high selectivity catalysts.

[0043] Table 1 : Parameters for low and high selectivity catalysts

Example 2

[0044] FIG. 2 A is an evaluation of a catalyst-in- shell side reactor design with cross flow (XCSA) for low selectivity catalyst with GHSV of 5631 1/hr. A comparison case is shown with a 2" tube OD (1.83"ID) conventional shell and tube reactor with catalyst-in-tube (STR) design. For low selectivity (LS) catalysts with range of 80 to 86% such as disclosed in Example 1 above, a catalyst-in-shell with cross flow (XCSA) design with 0.75" coolant tube with GHSV of 5631 1/hr will show advantages over the conventional shell and tube reactor with catalyst-in-tube (STR) design with 2" tube OD (with 1.83" ID). The STR case tube ID is such that the heat transfer area over catalyst volume ratio (φ) is 86 1/m. In this case, the XCSA design will show improved stability as shown by the larger calculated sensitivity index (SI), lower weight and lower pressure drop (ΔΡ) and even lower φ with XCSA reactor bed length between 6.7 and 1 1.7 m as shown in Figure 2 A. It is also apparent that the ΔΡ decreases with a lower bed length while in contrast the reactor weight increases with lower bed length.

[0045] FIG. 2B is an evaluation of catalyst-in-shell side reactor design with cross flow (XCSA) for low selectivity catalyst with GHSV of 7525 1/hr. A comparison case is shown with 2" tube OD (1.83"ID) conventional shell and tube reactor with catalyst-in-tube (STR) design. The reactor of Figure 2B will show a similar improvement as the reactor of Figure 2A by using an XCSA concept with a GHSV of 7531 1/hr. Table 2 shows the detailed calculation results from LS catalyst and it also shows that similar improvement may be expected to be achieved in XCSA designs over STR designs using 2" OD tubes (φ =86 1/m and 1.83"ID) by using different coolant tubes OD while keeping the catalyst bed volume and other operating conditions (coolant temperature, GHSV, production rate, inlet pressure, inlet gas temperature) similar to STR cases and with SI, EO outlet concentration, and selectivity similar to or better than those in STR cases. In addition, Table 2 also shows that significant improvement in the φ (and thus reactor weight) may be obtainable at various coolant tube OD (e.g. XCSA 2, 5 and 8) while still providing lower APs than that of STR cases. Note also that ΔΡ can be much lower than that in STR cases for various coolant tubes OD while still maintaining lower weight ratio (e.g. case XCSA 3, 6, and 9).

[0046] Table 2: Evaluation of catalyst-in- shell side reactor design with cross flow (XCSA) and comparison with conventional fixed bed reactor for lower selectivity catalyst for GHSV of 5631 1/hr.

[0047] FIG. 2C is a plot showing the prediction of heat transfer area to catalyst volume ratio φ as a function of reactor bed length for XCSA reactor design with various coolant tube diameters and various coolant temperatures yielding different work rates (work rate is indicated in legend in lbs/ft³-hr) for low selectivity catalyst. The ΔΡ, 1/SI and weight ratio with respect to STR cases (STR with 2" OD and D_ti = 1.83") are also plotted as a function of bed length. Figure 2C also shows the prediction that the XCSA design is advantageous over STR designs with

1/m for different coolant temperatures and hence different production rates. More importantly, Figure 2C also shows the prediction that the φ of XCSA design concept is always lower than or equal to c sTR of 86 1/m when the catalyst bed length is equal to or larger than 6.5 m. In addition, this is also predicted to be true for all coolant tubes OD of 0.6" to 1.5" and at various coolant temperatures. Note that Figure 2C also demonstrates that the XCSA design can provide lower weight, higher stability, and lower ΔΡ with bed length up to about an 11 m bed than that of the STR design.

[0048] FIG. 2D is plot showing prediction of heat transfer area to catalyst volume ratio φ as a function of reactor bed length for XCSA reactor design at different GHSV values for a low selectivity catalyst. The ΔΡ, 1/SI and weight ratio with STR cases (STR with 2" OD and D_ti = 1.83") are also plotted as a function of bed length. Figure 2D illustrates the predicted advantage of an XCSA design with lower φ for various GHSV values as compared to the STR case design with

1/m. As shown above in reference to Table 2, both the predicted reactor ΔΡ and stability are also advantageous over an STR design of up to 11 m bed length.

[0049] FIG. 2E is a plot showing a prediction of heat transfer area to catalyst volume ratio φ as a function of reactor bed length for XCSA reactor design with various coolant tube diameters and various coolant temperatures yielding different work rate (work rate is indicated in legend in lbs/ft³-hr) for a low selectivity catalyst. The predicted ΔΡ, 1/SI and weight ratio with an STR case (conventional reactor with D_ti = 0.84") are also plotted as a function of bed length. As is the case with Figures 2A through 2D, Figure 2E demonstrates the predicted advantages of an XCSA design concept with lower φ as compared to the STR design with tube OD of 0.84" and

1/m, at different coolant temperature and XCSA design coolant tube with OD of 0.75" and 1.5' with bed length in the range of 6.0 m to 12 m. This also illustrates that the XCSA design of Figure 2E is expected to show better expected stability, requires lower expected reactor weight and ΔΡ at the same operating conditions as STR with tube OD of 0.84" in the range of 6.5 m to 11 m bed length.

[0050] FIG. 2F is a plot showing a prediction of heat transfer area to catalyst volume ratio as a function of reactor bed length for XCSA reactor design with various tube diameters and various coolant temperatures yielding different work rate (work rate is indicated in legend in lbs/ft³ -hr) for a low selectivity catalyst. The expected ΔΡ, 1/SI and weight ratio with STR case (conventional reactor with tube ID of 1.5") are also plotted as a function of bed length. A similar trend to that indicated in Figures 2A through E is illustrated in Figure 2F for the XCSA design as compared to STR design case with tube ID of 1.5" and ( sTR- 105.0 1/m, at different XCSA coolant tube ODs and coolant temperatures. Figure 2F also depicts the expected XCSA advantageous bed length range of 5.8 m to 12.0 m. [0051] FIG. 2G is a plot showing a prediction of heat transfer area to catalyst volume ratio as a function of reactor bed length for an XCSA reactor design with various tube diameters and various coolant temperatures yielding different work rate (work rate is indicated in legend in lbs/ft³ -hr) for a low selectivity catalyst. The expected ΔΡ, 1/SI and weight ratio with STR case (conventional reactor with tube ID of 2.17") are also plotted as a function of bed length. A similar trend as seen in Figures 2A through 2F is also illustrated in Figure 2G for the XCSA design as compared to the STR case design with tube ID of 2.17" or

1/m, at different coolant tube OD and coolant temperature. Figure 2G also depicts the expected XCSA advantageous bed length range of 7.5 m to 11.5 m.

[0052] Finally, an expected overall net present value (NPV) improvement over the STR design of all the advantageous XCSA designs with various c sTR values is plotted against the bed length in Figure 2H. The expected overall NPV improvement expected from savings in operating cost (Operating ΔΡ) and capital cost (approximately proportional to reactor weight) as compared to STR case with tube OD of 2", shows a maximum along the bed length range. For lower reactor bed lengths, NPV savings from operating costs are expected to increase, due to lower pressure drop across the reactor, and higher bed length savings from capital investment are expected to be higher due to lower reactor weight. This gives rise to the highest expected NPV savings at an intermediate length range from 8 to 9.5 m. More importantly Figure 2H also shows that the expected NPV improvement of the XCSA design may begin to be realized for the case with c sTR of 186.4 1/m or tube OD of 0.84" in the STR design.

Example 3

[0053] FIG. 3A is a predicted evaluation of a catalyst-in-shell side reactor design with cross flow (XCSA) for a high selectivity catalyst with a GHSV of 6652 1/hr. The STR case is using a 2" tube OD (1.83"ID) for a conventional shell and tube reactor with catalyst-in-tube design. As depicted therein, for a high selectivity (HS) catalyst with range of 86 to 95%, the catalyst on the shell side with cross flow (XCSA) design with 0.75" coolant tube with GHSV 6652 1/hr is expected to show advantages over the conventional shell and tube reactor with catalyst-in-tube (STR) design with 2" tube OD (1.83"ID). The STR case has the same φ =86 1/m, as seen in Example 2. Figure 3 A depicts that the expected XCSA design requires lower φ than the STR design and also shows predicted improved stability, lower reactor weight and lower pressure drop (ΔΡ) with reactor bed length between 6 m and 9 m.

[0054] Similar improvements are also expected for a case with different GHSV as shown in Figure 3B. FIG. 3B is an evaluation of a predicted catalyst-in-shell side reactor design with cross flow (XCSA) for a high selectivity (HS) catalyst with a GHSV of 8500 1/hr. The STR case is with a 2" tube OD (1.83"ID) conventional shell and tube reactor with catalyst-in-tube design. Table 3 shows the detailed calculations expected results for HS catalyst and also shows that similar improvement can be expected to be achieved in an XCSA design over an STR design by using different coolant tube OD's while keeping catalyst volume and other operating conditions (coolant Temperature, GHSV, production rate, inlet pressure, inlet gas temperature) similar to the STR case. Table 3 also shows that significant improvement in the φ (thus reactor weight) is expected to be obtained at various coolant tube OD's (e.g. XCSA 2, 5 and 8) while still providing similar or lower ΔΡ than that of the STR case. Note that ΔΡ is expected to be much lower than that in STR case for various coolant tube OD while still maintaining lower weight ratio (e.g. case XCSA 1, 2, 6 and 7).

[0055] Table 3: Catalyst-in-shell side reactor designs with cross flow (XCSA) and comparison with conventional fixed bed reactor for high selectivity catalyst with GHSV of 6652 1/hr.

[0056] FIG. 3C is a plot showing a prediction of heat transfer area to catalyst volume ratio for high selectivity catalyst as a function of reactor bed length for an XCSA reactor design with various tubes. Work rate is indicated in legend in lbs/ft³ -hr. The expected ΔΡ, 1/SI and weight ratio with STR case tube OD's of 2" (ID of 1.83") are also plotted as a function of bed length. Figure 3C shows that various XCSA designs are expected to be advantageous over an STR design with

1/m for different coolant temperatures and hence different production rates. Figure 3C also shows that the φ of an XCSA design concept is always lower than or equal to c sTR of 86 1/m when bed length is equal to or larger than 6.2 m and this is valid for all coolant tubes OD of 0.75" to 1.5" and at various coolant temperatures. Figure 3C also demonstrates that an XCSA design concept is expected to have lower weight, stability, and ΔΡ up to a bed length of about 9.5 m as compared to an STR design. As shown above, both the predicted reactor ΔΡ and stability are also advantageous over an STR design up to 9.5 m bed length.

[0057] FIG. 3D is a plot showing a prediction of heat transfer area to catalyst volume ratio as function of reactor bed length for an XCSA reactor design at different GHSV values. The expected ΔΡ, 1/SI and weight ratios with the STR case (conventional reactor with D_ti =1.83") are also plotted as a function of bed length for a high selectivity catalyst. Figure 3D illustrates the expected advantage of an XCSA design with lower φ for various GHSV values as compared to an STR case design with

1/M. As shown above, both the expected reactor ΔΡ and stability are also advantageous over an STR design up to 9.5 M bed length.

[0058] FIG. 3E is a plot showing a prediction of heat transfer area to catalyst volume ratio as function of reactor bed length for an XCSA reactor design with various coolant temperatures yielding different work rate for a high selectivity catalyst. Work rate is indicated in legend in lbs/ft³ -hr. The expected ΔΡ, 1/SI and weight ratio with an STR case (conventional reactor with tube ID of 0.84") are also plotted as a function of bed length. Figure 3E is similar to Figures 3A through 3D, and demonstrates the predicted advantages of an XCSA design concept with lower φ as compared to an STR case design with tube OD of 0.84" and

1/M, at different coolant temperature and design coolant tubes with OD of 0.75" with bed length in the range of 7.5 M to 9.0 M. Figure 3E also illustrates that the XCSA design shows better expected stability, requires lower reactor weight and ΔΡ at the same operating conditions as an STR with tube OD of 0.84" in the range of 7.5 to 9.8 M bed length.

[0059] FIG. 3F is a plot showing a prediction of heat transfer area to catalyst volume ratio as a function of reactor bed length for an XCSA reactor design with various coolant

temperatures yielding different work rates for a high selectivity catalyst. Work rate is indicated in legend in lbs/ft³-hr. The expected ΔΡ, 1/SI and weight ratio with the STR case (conventional reactor with tube ID of 1.5") are also plotted as a function of bed length.

Figure 3F is similar to Figures 3A through 3E, and sets forth the predicted advantages of an XCSA design as compared to an STR design case with tube ID of 1.5" and

1/M, at different XCSA coolant temperatures. Figure 3F also depicts the XCSA is expected to have an advantageous bed length range of 7.8 M to 9.0 M.

[0060] FIG. 3G is a plot of predicted NPV savings of a feasible XCSA design with coolant tube OD of 0.75" as a function of bed length, as compared to an STR case with various heat transfer area ratio to catalyst volume (or tube ID) for a high selectivity catalyst. The expected overall net present value (NPV) improvement over the STR design of all the advantageous XCSA designs with various c sTR values are plotted against the bed length in Figure 3G for the high selectivity catalyst system. The expected overall NPV improvement coming from savings in operating cost (proportional to the operating ΔΡ) and capital cost (proportional to the reactor weight) as compared to base case STR with tube OD of 2", shows a maximum along the bed length range. For lower reactor bed length, expected NPV savings from operating costs are higher as seen in Figure 3G, and for higher bed length, predicted savings from capital investments are higher due to lower reactor weight, as seen in Figure 3F. This gives rise to a highest expected NPV at an intermediate length range from 7.0 to 8.5 m. More importantly Figure 3G also shows that the NPV improvement of the XCSA design may be expected to be realized for the case with c sTR of 186.4 1/M or tube OD of 0.84" in the STR design.

Example 4

[0061] This example illustrates the predicted impact of a reactor catalyst bed with varying areas in the direction of process flow. The ratio of the absolute difference between outlet and inlet area over the catalyst bed length A_L of less than 1.3 M indicates where the reactor could be operated with sufficient stability. For a tubular type of reactor this can be represented with a truncated cone shape catalyst bed with an angle of 9° as shown in Figure 4. This 9° angle represents the predicted expansion in the shell and tubes such that heat transfer area to catalyst volume is maintained at 67 (1/M). When A_L is larger than 1.3 M as is the case in the radial flow reactor as shown in Table 5, the reaction is predicted to run away since the flow rate would decrease with bed length and reduce the heat transfer rate. [0062] Table 5. Design variables of radial flow design and variable area design at constant catalyst volume with coolant tube OD of 0.75" and φ = 67 1/M.

Example 5

[0063] Figure 5A shows the predicted feasible catalyst-in-shell design with reactant gas flowing parallel (CSA) to the heat transfer surface area with LS catalyst as seen Example 1) with a coolant tube OD of 0.75". This case is compared with a 2" tube OD (1.83"ID) conventional shell and tube reactor with catalyst-in-tube (STR) design. The heat transfer area to catalyst volume required for the CSA design is expected to be the same as the STR case (φ= 86 1/M). The cross flow configuration (XCSA) is expected to provide better heat transfer than the CSA design with lower φ values (lowest φ value required for XCSA design is 22% lower than the STR case). The heat transfer coefficient for the XCSA design with cross flow configuration is predicted to be almost twice that of the CSA design. The wider feasible catalyst-in-shell design window for the XCSA design is expected to be achieved with a bed length in the range of 6.7 - 11 M as compared to the CSA design with a bed length range of 9-11 M. This may be seen in a comparison of Figure 2A and Figure 5 A. The expected minimum ΔΡ for the XCSA design is 80% lower than the STR compared to expected minimum ΔΡ for the CSA design, which is only 50% lower than the STR case. The predicted minimum reactor weight for the feasible XCSA design is 42% lower than the STR case as compared to the predicted minimum reactor weight for CSA design, which is 31%> lower than the STR case. Similarly, Figure 5B also shows the expected advantages of the CSA design with coolant tube OD of 1" as compared to the STR case. The overall expected performance of the XCSA design is better than the CSA design with lower φ values, lower ΔΡ, lower reactor weight and better reactor stability. Example 6

[0064] Figure 6A shows the predicted feasible catalyst-in-shell design with reactant gas flowing parallel (CSA) to the heat transfer surface area with HS catalyst as seen in example 1) with a coolant tube OD of 0.75". The heat transfer area to catalyst volume required for the CSA design is the same as the STR case (φ = 86 1/M). The cross flow configuration of the predicted XCSA provides better heat transfer than a CSA design with lower φ values. Note that the lowest expected φ value required for the XCSA design is 20% lower than the STR case. The heat transfer coefficient for an XCSA design with cross flow configuration is predicted to be almost twice that of a CSA design. The wider feasible catalyst-in-shell design window for an XCSA design may be achieved with a bed length in the range of 6.0 - 9 M as compared to a CSA design with a bed length range of 7.6-8.8 M. This can be seen in a comparison of Figure 3 A and Figure 6A. The expected minimum ΔΡ for an XCSA design is 50% lower than the STR compared to the predicted minimum ΔΡ for the CSA design, which is only 37% lower than the STR case. The expected minimum reactor weight for the feasible XCSA design is 37%> lower than the STR case as compared to predicted minimum reactor weight for CSA design, which is 27% lower than the STR case. The STR case depicted is with a 2" tube OD (1.83"ID) conventional shell and tube reactor with catalyst-in-tube (STR) design.

[0065] FIG. 6B is an evaluation of a predicted catalyst-in-shell side axial flow design with flow parallel to the coolant carrier (CSA) for high selectivity catalyst as compared to the STR case. As is the case with Figure 6A, Figure 6B shows the expected advantages of a CSA design with coolant tube OD of 1 " as compared to the STR case. The overall performance of the XCSA design is predicted to be better than the CSA design with lower φ values, lower ΔΡ, lower reactor weight and better reactor stability.

Example 7

[0066] The expected porosity and catalyst bed density effect on low selectivity EO catalyst system with XCSA design is shown in Figure 7. The porosity (ε) is varied from 0.4 to 0.48 as compared to typical value of 0.44. Figure 7 shows that for all the porosity ranges and the corresponding catalyst bed density ranges, the XCSA design is expected to always perform better than the corresponding STR case with lower ΔΡ, lower reactor weight, and better stability. The predicted preferred range of catalyst bed porosity is within 0.43-0.45.

Example 8

[0067] The predicted effect of porosity and catalyst bed density on a high selectivity EO catalyst system with an XCSA design is shown in Figure 8. The porosity (ε) is varied from 0.4 to 0.48 as compared to a typical value of 0.435. Figure 8 shows that for all the porosity ranges and the corresponding catalyst bed density ranges, the XCSA design is always predicted to perform better than the conventional STR case with lower ΔΡ, lower reactor weight, and better reactor stability. The expected preferred range of catalyst bed porosity is within 0.42-0.44.

[0068] Those skilled in the art recognize that the words used in this specification are words of description and not words of limitation. Many variations and modifications will be apparent to those skilled in the art upon a reading of this application without departing from the scope and sprit of the invention as set forth in the appended claims.

Claims

What is claimed as new and desired to be protected by Letters Patent of the United States is:

1. A method for producing gaseous ethylene oxide product from partial oxidation of hydrocarbon using a heterogeneous catalyst in a fixed bed enclosed within a reaction vessel shell, comprising:

introducing a sufficient amount of gaseous ethylene, oxygen, ballast gas, and catalyst promoter into a reaction vessel having an inflow and an outflow, said vessel having a shell with a length and a width to define a reactor vessel volume; said reactor vessel volume defining a catalyst bed shape; said catalyst bed having a length such that an out flow area and an in flow area over said catalyst bed length in between the reactor outflow and inflow has an absolute ratio difference less than or equal to about 1.3 M anywhere in the reactor bed; said catalyst bed defining a process side and having a selectivity greater than about 80%, said catalyst bed length less than said shell length and having a volume less than said reactor shell volume;

circulating a heat transfer fluid through at least one coolant enclosure heat exchanger in said shell interior to define a coolant side of the process; said coolant enclosure equipped with an outside surface and an inside surface; said coolant enclosure outside surface in contact with said catalyst bed and having an inlet and an outlet for the flow of coolant fluid therethrough; said coolant enclosure further defining a cooling surface area with a coolant flow cross sectional area ratio to cooling surface area less than about 1 ; said coolant side having a higher than said process side; and

flowing said gaseous ethylene, oxygen, ballast gas and catalyst promoter over said catalyst bed and through a fixed catalyst bed outlet zone in said reactor vessel; said zone configured with average residence time less than or equal to 4 seconds of said gaseous ethylene, oxygen, ballast and catalyst promoter flow from the catalyst bed to the heat exchanger to quench said process to create gaseous ethylene oxide product and flowing said gaseous ethylene oxide product through said reactor vessel outlet.

2. The method of claim 1, wherein the flow of gaseous ethylene, oxygen, ballast gas, and catalyst promoters through the catalyst bed is crosswise to said heat transfer fluid flow through said coolant enclosure.

3. The method of claim 1, wherein said inflow for ethylene, oxygen, ballast gas, and catalyst promoters, and said outflow for ethylene oxide are configured to produce an exit gas velocity of at least about 5 ft/s upon exiting the fixed catalyst bed within the reactor vessel outlet before entering an outlet pipe.

4. The method of claim 1, wherein said total coolant surface area ratio to catalyst bed volume is less than or equal to about 187 (1/M).

5. The method of claim 1, further including removing ethylene oxide containing effluent from the reaction catalyst bed heating elements and collecting said effluent in a head chamber to minimize back mixing before quenching said oxidation reactions.

6. The method of claim 1, wherein said heat exchange fluid is boiling water in said coolant enclosure at a pressure up to about 750 psig to maintain temperature in the catalyst bed up to about 270°C.

7. The method of claim 1, wherein the shell has an interior pressure ingress of said gaseous ethylene, oxygen and gaseous product egress of less than about 350 psig.

8. The method of 1, wherein the coolant enclosure cross sectional area is at least equivalent to one tube having an inner diameter in the range of from up to about 0.2" to about 1.8".

9. The method of claim 1, wherein said inlet for ethylene and said outlet for ethylene oxide containing effluent product are configured to produce an gas velocity exiting the fixed bed in the range of from about 15 ft/s to about 25 ft/s.

10. The method of claim 1, wherein said inlet gas comprising ethylene, oxygen, ballast gas and promoters, and said ethylene oxide effluent products outlet have an absolute difference ratio of the inlet and outlet flow area over said catalyst bed length of from about 0.8 M to about 1.3 M and more preferably, from about 0.9 M to about 1.2 M.

12. The method of claim 1, wherein the flow of gaseous ethylene, oxygen, ballast gases and catalyst promoter through the catalyst bed is parallel to said heat transfer fluid flow through said coolant enclosure.

13. The method of claim 1, wherein said catalyst is comprised of pills having a diffusion length of from about 0.02 inches to about 0.07 inches.

14. The method of claim 1, wherein said catalyst bed length is larger than about

9.5 M.

15. The method of claim 1, wherein said catalyst bed length is larger than about

8.5 M.

16. The method of claim 1, wherein said catalyst bed has a porosity in the range of from about 0.38 to about 0.49.

17. A reaction vessel for production of gaseous alkylene oxide from partial oxidation of hydrocarbon, catalyst promoter and ballast gas using a high efficiency heterogeneous catalyst in a fixed bed enclosed within a reaction vessel shell, comprising: a shell having an inflow and an outflow; said shell with a length and width to define a reactor vessel interior volume;

a catalyst bed shape defined by said reactor volume, said catalyst bed having a length such that an out flow area and an in flow area over said catalyst bed length in between the reactor outflow and inflow has an absolute ratio difference less than or equal to about 1.3 M anywhere in the catalyst bed; said catalyst bed including a catalyst having a selectivity greater than about 80%, said catalyst bed defining a process side and having a length less than said shell length and a volume less than said shell volume; a fixed bed outlet zone configured with an average residence time less than or equal to 4 seconds of a gaseous flow through said catalyst bed to a heat exchanger to quench the reactions; said catalyst bed forming a process side of said reaction vessel; and

at least one coolant enclosure heat exchanger in said shell interior volume to define a coolant side of the reaction vessel, said coolant enclosure having an outside surface and an inside surface to define a heat transfer fluid flow path; said coolant enclosure outside surface in contact with said catalyst, said coolant enclosure having an inlet and an outlet for flow of heat transfer fluid through said coolant enclosure inner surface; said coolant enclosure further defining a cooling surface area with the coolant flow cross sectional area ratio to a cooling surface area much less than 1 ; said coolant side having a pressure greater than or equal to said process side.

18. The reaction vessel of claim 19, wherein said inlet for gaseous alkylene, oxygen, catalyst promoter and ballast gas and said outlet for gaseous alkylene oxide are configured to produce a gas velocity exiting the fixed bed in the range of from about 5 ft/s to about 25 ft/s.